Attribution Modeling
Also: Marketing Attribution, Multi-Touch Attribution, MTA, Conversion Attribution
Attribution modeling is the method of assigning credit for a conversion across the marketing touchpoints a customer interacted with before buying or signing up.
What It Is
Attribution modeling is the practice of distributing credit for a conversion (a sale, signup, lead, or other goal) across the various marketing touchpoints a customer encountered on their path. Because most buyers interact with several channels (search ads, email, social, organic, referral) before converting, attribution answers a core question: which touchpoints actually drove the result, and how much credit does each deserve?
Why it matters
Marketing budgets are finite, and channels rarely work in isolation. Without attribution, teams tend to over-reward the last click and under-invest in awareness channels that started the journey. Good attribution helps you:
- Allocate budget toward channels that genuinely influence outcomes.
- Measure true return on ad spend (ROAS) per channel.
- Justify spend to finance with defensible logic.
- Detect wasteful or redundant touchpoints.
Common Models
- First-touch: 100% credit to the first interaction. Good for measuring demand generation.
- Last-touch: 100% credit to the final interaction before conversion. Simple but ignores the funnel.
- Linear: equal credit across all touchpoints.
- Time-decay: more credit to touchpoints closer to conversion.
- Position-based (U-shaped): heavy credit to first and last, the rest split among the middle.
- Data-driven (algorithmic): uses statistical or machine learning methods to assign credit based on observed patterns, rather than fixed rules.
How it is used in practice
Teams configure attribution in analytics or marketing platforms, define a lookback window (for example, 30 or 90 days), and compare model outputs. Many run several models side by side, since each tells a different story. Privacy changes (cookie loss, consent rules) increasingly push teams toward modeled and aggregated approaches.
Concrete Example
A customer sees a LinkedIn ad, later clicks a Google search ad, then opens a retargeting email and buys a $600 subscription.
- Last-touch: email gets all $600.
- Linear: each of the three gets $200.
- Position-based: LinkedIn $240, email $240, search $120.
The same conversion produces very different channel scorecards, which is why choosing and documenting your model matters.
Frequently asked questions
What is attribution modeling in marketing?
Attribution modeling is the method of distributing credit for a conversion (a sale, signup, or lead) across the marketing touchpoints a customer encountered before converting. Since most buyers touch search ads, email, social and organic before they act, attribution answers which touchpoints actually drove the result and how much credit each one deserves.
What is the difference between first-touch and last-touch attribution?
First-touch gives 100% of the credit to the initial interaction, last-touch gives 100% to the final one before conversion. First-touch is useful for judging demand generation, last-touch is simple but ignores everything that happened earlier in the funnel. Teams that only look at last-touch systematically under-invest in awareness channels.
Which attribution model should I pick for my company?
There is no single correct model, which is why many teams run several side by side and compare the outputs. Position-based works when the first and last touch matter most, time-decay when the sales cycle is short, data-driven when you have enough volume to let statistical methods assign credit from observed patterns. What matters more than the choice itself is documenting the model so everyone reads the same scorecard.
How much can results change from one attribution model to another?
Enough to reverse a budget decision. Take a $600 subscription where the customer saw a LinkedIn ad, clicked a Google search ad, then converted through a retargeting email: last-touch credits the full $600 to email, linear gives $200 to each channel, position-based gives LinkedIn $240, email $240 and search $120. Same conversion, three completely different channel scorecards.
What is a lookback window and how does it affect attribution?
The lookback window is the period before a conversion during which touchpoints are still counted, typically 30 or 90 days. A short window drops early interactions and inflates the channels closest to the sale; a long one captures the full journey but may credit contacts that had little real influence. Set it against your actual sales cycle length, not by default.